pdfdsf435/tinygpt-v283-data download history
pdfdsf435/tinygpt-v283-data is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 1,419 times (1,419 in the last 7 days), and 1,419 times in total. It ranks #14,842 among datasets by monthly downloads.
tinygpt-v283 compiled data (v11) Pre-tokenized SmolLM2-135M (vocab 49152) uint16 token shards. Mixture fineweb-edu-dedup: 79% cosmopedia-v2: 10% wikipedia: 6% smoltalk-instruct: 5% Total pretrain tokens: 6.40B Total SFT examples: 600001 Filters (v11) E